Potential of the test-negative design for measuring influenza vaccine effectiveness: a systematic review

Abstract

Background: The test-negative design is a variant of the case–control study being increasingly used to study influenza vaccine effectiveness (VE). In these studies, patients with influenza-like illness are tested for influenza. Vaccine coverage is compared between those testing positive versus those testing negative to estimate VE. Objectives: We reviewed features in the design, analysis and reporting of 85 published test-negative studies. Data sources: Studies were identified from PubMed, reference lists and email updates. Study eligibility: All studies using the test-negative design reporting end-of-season estimates were included. Study appraisal: Design features that may affect the validity and comparability of reported estimates were reviewed, including setting, study period, source population, case definition, exposure and outcome ascertainment and statistical model. Results: There was considerable variation in the analytic approach, with 68 unique statistical models identified among the studies. Conclusion: Harmonization of analytic approaches may improve the potential for pooling VE estimates.

Keywords::

• case–control
• influenza
• public health
• test-negative study
• vaccine effectiveness

Financial & competing interests disclosure

BJ Cowling has received research funding from MedImmune Inc. and Sanofi Pasteur for influenza vaccine efficacy and effectiveness studies, and has consulted for Crucell NV on pharmaceutical options for influenza control. This work has received financial support from the Area of Excellence Scheme of the Hong Kong University Grants Committee (grant no. AoE/M-12/06) and the Harvard Center for Communicable Disease Dynamics from the National Institute of General Medical Sciences (grant no. U54 GM088558). The WHO Collaborating Centre for Reference and Research on Influenza is funded by the Australian Government Department of Health. The funding bodies were not involved in the collection, analysis and interpretation of data, the writing of the article or the decision to submit it for publication. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Key issues

• The test-negative design is growing in popularity as a feasible and efficient approach to estimation of influenza vaccination effectiveness.

• These studies often take advantage of existing surveillance systems that routinely monitor influenza-like illness in hospitals or the community. However, specific research studies have also been conducted using this design.

• Although all these studies used the same basic design, there is tremendous variation in restriction criteria and the specification of variables included in logistic regression models.

• We show how some of the design features commonly employed to reduce bias may have little impact on estimates. For example, restricting patients to only those presenting soon after illness onset may only minimally reduce estimates. In contrast, many studies appear to rely on patient recall or general practice records to ascertain vaccination status, and this may be a greater source of measurement error.

• Because of the unusual emphasis placed on variations of the odds ratio estimate in these types of studies it may be prudent for a common minimal study design to be employed to facilitate comparison across studies.

Notes